Glioblastoma multiforme (GBM) is almost universally fatal. The discovery of tumor-initiating cells with the capacity to self-renew, sometimes termed "cancer stem cells", has created tremendous enthusiasm for the development of new avenues of therapy. These cells utilize familiar pathways for their proliferation, such as the PI3 Kinase pathway. Despite the hope raised by the discovery of brain tumor stem cell-like cells, numerous obstacles lie in the path of therapeutic development. One complication is that these cells have significant resistance to conventional therapies and to inhibition of pathways. Another is that there are differences amongst brain tumor stem-like cells that are present in the tumors of different patients and probably amongst the multiple types of such cells within individual patient's tumors. The goals of this study are to critically examine brain tumor stem cell-like cell biology in order to develop the means to attack them and to overcome their mechanisms of resistance. First, we will determine the genetic mechanisms underlying the dependence of GBM stem cell-like cells on the PI3K pathway. We will use a limited sequencing approach to determine the mutational spectrum of the stem cell-like cells derived from different patients and that of different clonal lines derived from individual patients with the goal of establishing the fundamental genotype-phenotype relationships in these biologically important sets of cells. We will test whether those cells with mutations that activate the PI3K pathway are more dependent on this pathway for proliferation and tumorigenesis than the stem cell-like cells with mutations in other pathways. We will next assess the role of the PI3K pathway in mediating the enhanced resistance to radiation observed in brain tumor stem cell-like cells.To identify potential mechanisms of this radioresistance, we will determine if PTEN modulates autophagy and also whether the PI3K pathway promotes survival through an antioxidant response following ionizing radiation in the same cells. We will then explore mechanisms of chemoresistance in GBM stem cell-like cells. We will use cell culture, in vivo assays and microfluidics-based immunocytochemical analysis (MIC) to determine whether rapamycin selects for stem cell-like cells with enhanced tumorigenicity and pathway activation. We will also determine whether resistance to rapamycin treatment can be overcome through inhibition of hyperactivated pathways. Then, we will identify pathways of resistance based on a completed phosphoproteomic screen to discover proteins that are phosphorylated or dephosphorylated during the development of rapamycin resistance. We will determine the potential role of the proteins identified by this screen in the development of resistance. Finally, we will use MIC to study the mechanisms of resistance to the EGF receptor inhibitor, erlotinib, a drug that has been proposed as a molecularly targeted therapy for a subset of GBM patients. We will determine whether erlotinib selects against EGFRvIII-positive calls, and whether these resistant cells have cancer stem cell properties. Furthermore, we will identify the means to overcome this resistance.